Search

Realism about computation is the view that whether or not a particular physical system is performing a particular computation is at least sometimes a mindindependent feature of reality. The caveat 'at least sometimes' is necessary here because a realist about computation need not believe that all instances of computation should be realistically construed. The computational theory of mind presupposes realism about computation. If whether or not the human nervous system implements particular computations is not a natural fact about the world that is independent of whether we represent it as doing so, then the computational theory of mind fails to naturalise the mind. Realism about computation is also presupposed by attempts to use computational principles such as Landauer's Principle to dispel Maxwell's Demon. Realism about computation has been challenged by Hilary Putnam and John Searle among others. Various arguments have been put forward purporting to show that any physical system of sufficient complexity trivially implements all computations. Ladyman et al. (2007) offer a precisification and general proof Landauer's Principle. In order to do this they present an analysis of what it is for a physical process to implement a logical transformation. In this paper, their analysis is explained and its implications for realism about computation and the use of Landauer's Principle in foundational debates is assessed. [ABSTRACT FROM AUTHOR]

In this paper we prove the Uniformity Principle for Σ-definability over the real numbers extended by open predicates. Using this principle we show that if we have a ΣK-formula, i.e. a formula with quantifier alternations where universal quantifiers are bounded by computable compact sets, then we can eliminate all universal quantifiers obtaining a Σ-formula equivalent to the initial one. We also illustrate how the Uniformity Principle can be employed for reasoning about computability over continuous data in an elegant way. [ABSTRACT FROM AUTHOR]

We have mentioned the MDL principle on several occasions somewhat loosely as the principle that calls for finding the model and model class with which the data together with the model and model class, respectively, can be encoded with the shortest code length. Actually to apply the principle we must distinguish between two types of models — those for data compression and others for general statistical purposes such as prediction. In data compression, we apply the models to the same data from which the models are determined. Hence these models need not have any predictive power; and, in fact, to get the shortest code length we do not even need to fit models in the class considered, say, $$ \mathcal{M}_\gamma $$γ. This is because the universal NML model gives a code length, which we called the stochastic complexity and which we consider to be the shortest for all intents and purposes. [ABSTRACT FROM AUTHOR]

Bloat control and generalization pressure are very important issues in the design of Pittsburgh Approach Learning Classifier Systems (LCS), in order to achieve simple and accurate solutions in a reasonable time. In this paper we propose a method to achieve these objectives based on the Minimum Description Length (MDL) principle. This principle is a metric which combines in a smart way the accuracy and the complexity of a theory (rule set , instance set, etc.). An extensive comparison with our previous generalization pressure method across several domains and using two knowledge representations has been done. The test show that the MDL based size control method is a good and robust choice. [ABSTRACT FROM AUTHOR]

One can interpret the No Probabilities for Acts-Principle, namely that any adequate quantitative decision model must in no way contain subjective probabilities for actions in two ways: it can either refer to actions that are performable now and extend into the future or it can refer to actions that are not performable now, but will be in the future. In this paper, I will show that the former is the better interpretation of the principle. [ABSTRACT FROM AUTHOR]

It has been well known that edge filters in the visual system can be generated by the InfoMax principle. In this paper, the "InfoMin" principle is proposed, which asserts that the information through some neighboring signals on a two-dimensional mapping must be minimized. It is shown that the standard Comon's ICA can be derived from the combination of the InfoMax principle for the whole signals and the InfoMin one for each signal under a linear model with sufficiently large noise. It is also shown that the InfoMin principle for the signals within neighboring areas can generate a topographic mapping in the same way as in topographic ICA. [ABSTRACT FROM AUTHOR]

This paper describes characteristics of adaptive vector quantization according to the equinumber principle. Three methods of adaptive vector quantization are presented with the objective of avoiding the initial dependency of reference vectors. The present approaches which have output units without neighboring relations equalize the numbers of inputs in a partition space. The first approach is a creation method which sequentially creates output units to reach a predetermined number of neurons founded on the equinumber principle in the learning process. The second is a reduction method which sequentially deletes output units to reach a prespecified number. The third is an unification method of the creation and reduction methods, which deletes units after creating under the predetermined number. Experimental results show the properties of the present techniques. [ABSTRACT FROM AUTHOR]

In this paper a new learning algorithm, Lever Training Machine (LTM), is presented for binary classification. LTM is a supervised learning algorithm and its main idea is inspired from a physics principle: Lever Principle. Figuratively, LTM involves rolling a hyper-plane around the convex hull of the target training set, and using the equilibrium position of the hyper-plane to define a decision surfaces. In theory, the optimal goal of LTM is to maximize the correct rejection rate. If the distribution of target set is convex, a set of such decision surfaces can be trained for exact discrimination without false alarm. Two mathematic experiments and the practical application of face detection confirm that LTM is an effective learning algorithm. [ABSTRACT FROM AUTHOR]

In order to provide effective support to the principle of least privilege, considering the limitation of traditional privilege mechanisms, this paper proposes a new privilege control model called State-Based Privilege Control (SBPC) and presents the design and implementation of a prototype system for SBPC called Controlled Privilege Framework (CPF) on the Linux operating system platform. SBPC decomposes the time space of a process' lifetime into a series of privilege states according to activities of the process and its need for special permissions. The privilege state is closely related to the application logic of a process. It is the privilege state transfer event that stimulates a process to transfer from one privilege state into another one. For a specified process, there is a specific set of privileges corresponding to every privilege state of the process. With the implementation of CPF, experiment results show that fine-grain and automatic privilege control can be exercised transparently to traditional applications, threats of intrusion to a system can be reduced greatly, and support to the principle of least privilege can therefore be achieved effectively. [ABSTRACT FROM AUTHOR]

The principle of polyrepresentation, proposed more than 10 years ago, offers a holistic theoretical framework for handling multiple contexts in Information Retrieval (IR), and allows integration of representation and matching of both documents as well as the information seeker's information need in context. Relatively few empirically based studies have, however, applied the principle explicitly for IR purposes. This paper examines the principle of polyrepresentation, and analyses the practical implications of applying it to multiple contexts in best match IR research. It is concluded that the principle is inherently Boolean in its foundation in spite of its intentions to be applicable to both exact and best match IR. This may constitute a major obstacle for the application of the principle in main stream IR and information seeking research. A polyrepresentation continuum is proposed as an illustration of this problem, and as a model for developing the principle towards greater practical applicability. [ABSTRACT FROM AUTHOR]

The present fuzzy arithmetic is based on possibilistic extension-principle of Zadeh. Fuzzy arithmetic is necessary to accomplish calculation in Computing with Words [6]. However, arithmetic operations carried out with the standard, possibilistic extension-principle often yield results which have small or no practical value. Sometimes the results are paradoxical ones. An example of such case is presented in the paper and also a new cardinality extension-principle in a version that enables regard for all additional constraints resulting from the context of the problem to be solved. The principle can be used only to fuzzy sets of numbers of probabilistic character which often occur in practical problems. [ABSTRACT FROM AUTHOR]

We study, in this work, the maximum principle for the Beltrami color flow and the stability of the flow's numerical approximation by finite difference schemes. We discuss, in the continuous case, the theoretical properties of this system and prove the maximum principle in the strong and the weak formulations. In the discrete case, all the second order explicit schemes, that are currently used, violate, in general, the maximum principle. For these schemes we give a theoretical stability proof, accompanied by several numerical examples. [ABSTRACT FROM AUTHOR]

SHAI is developing a comprehensive Operator Machine Interface Assistant (OMIA) system. The system will assist operators learn the new common-cockpit MH-60R and MH-60S helicopters in an increasingly broad variety of mission tasks and analyses, using the wide assortment of sensor, navigation, and computational resources available. We are employing a principle hierarchy based intelligent tutoring system, to teach the Common Cockpit machine interface during simulated mission execution. The Intelligent tutoring system utilizes a SHAI developed ITS authoring environment that has been developed and enhanced for various other ITS projects. This paper describes the OMIA system, emphasizing the authoring tool's ability to develop visually the principle hierarchy and the evaluation finite state machines necessary to evaluate the student. In addition the decision tree enhancement to the authoring tool developed as part of this OMIA project, is described. [ABSTRACT FROM AUTHOR]

In this paper an application of the Complexity Approximation Principle to the non-linear regression is suggested. We combine this principle with the approximation of the complexity of a real-valued vector parameter proposed by Rissanen and thus derive a method for the choice of parameters in the non-linear regression. [ABSTRACT FROM AUTHOR]

Bringing software process change to an organisation is a real challenge. The authors have shown a sample attempt to carry out a process change and then reflected on its results and context. The present reflection points to a need for a set of principles and practices that would support the fragile process of introducing agility. For a start, the authors propose the Empirical Evidence principle exemplified using DICE® and the practice of Joint Engagement of the management and the developers. Both are results of a real-world process change case study in Poland. [ABSTRACT FROM AUTHOR]

A key issue in statistics and machine learning is to automatically select the "right" model complexity, e.g. the number of neighbors to be averaged over in k nearest neighbor (kNN) regression or the polynomial degree in regression with polynomials. We suggest a novel principle (LoRP) for model selection in regression and classification. It is based on the loss rank, which counts how many other (fictitious) data would be fitted better. LoRP selects the model that has minimal loss rank. Unlike most penalized maximum likelihood variants (AIC,BIC,MDL), LoRP only depends on the regression functions and the loss function. It works without a stochastic noise model, and is directly applicable to any non-parametric regressor, like kNN. [ABSTRACT FROM AUTHOR]

Automated segmentation of MR images is a difficult problem due to the complexity of the images. Up to now, several approaches have been proposed based on spectral characteristics of MR images, but they are sensitive to the noise contained in the MR images. In this paper, we propose a robust method for noisy MR image segmentation. We use region-based features for a robust segmentation and use principle component analysis (PCA) to reduce the large dimensionality of feature space. Experimental results show that the proposed method is very tolerant to the noise and the segmentation performance is significantly improved. [ABSTRACT FROM AUTHOR]

In this paper, we try to develop an analytical framework for classifying human basic emotions. We try to find out what are the major components of each facial expression, what are the patterns that distinguish them from one another. We applied widely used pattern recognition technique-principle component analysis to characterize the feature point displacements of each basic human facial expression for each individual in the existing database. For faces not existent in the database, so called "novel face" in our experiment, we will first find the face in the database that has most likely neutral face to this individual, and base on an assumption that are widely accepted in cognitive science, we will classify this novel face to the category where the most similar one belongs, and classifying his/her facial expression using the so called "expression model" of the most similar individual. This kind of approach has never be exploited before, then we will examine its robustness in our experiment. [ABSTRACT FROM AUTHOR]

The statistical learning theory has formulated the Structural Risk Minimization (SRM) principle, based upon the functional form of risk bound on the generalization performance of a learning machine. This paper addresses the application of this formula, which is equivalent to a complexity penalty, to model selection tasks for decision trees, whereas the quantization of the machine capacity for decision trees is estimated using an empirical approach. Experimental results show that, for either classification or regression problems, this novel strategy of decision tree pruning performs better than alternative methods. We name classification and regression trees pruned by virtue of this methodology as Statistical Learning Intelligent Trees (SLIT). [ABSTRACT FROM AUTHOR]

This paper introduces a Mutual Information Independence Model (MIIM) and proposes a feature relaxation principle to resolve the data sparseness problem in MIIM-based named entity recognition via hierarchical features. In this way, a named entity recognition system with better performance and better portability can be achieved. Evaluation of our system on MUC-6 and MUC-7 English named entity tasks achieves F-measures of 96.1% and 93.7% respectively. It also shows that 20K words of training data would have given the performance of 90 percent with the hierarchical structure in the features compared with 30K words without the hierarchical structure in the features. This suggests that the hierarchical features provide a potential for much better portability. [ABSTRACT FROM AUTHOR]

A new method for digital image scrambling transformation, namely Magic Cube Transformation, was proposed and its principle was introduced. Combined with the Logistic mapping in non-linear dynamics system, an image encrypting and decrypting algorithm based on Magic Cube Transformation was designed. A natural chaotic sequence was created with the key. The image matrix was transformed with this chaotic sequence, using the Magic Cube Transformation. Then an encrypted image was resulted. The decrypting operation was the reverse process of encryption. The experimental results indicate that the proposed algorithm can get satisfying effect. Finally, the characteristics of the algorithm were summarized and the aspects of the subsequent work were prospected. [ABSTRACT FROM AUTHOR]

Although every terminating loop has a ranking function, not every loop has a ranking function of a restricted form, such as a lexicographic tuple of polynomials over program variables. The polyranking principle is proposed as a generalization of polynomial ranking for analyzing termination of loops. We define lexicographic polyranking functions in the context of loops with parallel transitions consisting of polynomial assertions, including inequalities, over primed and unprimed variables. Next, we address synthesis of these functions with a complete and automatic method for synthesizing lexicographic linear polyranking functions with supporting linear invariants over linear loops. [ABSTRACT FROM AUTHOR]

We apply the Clonal Selection principle of the human immune system to solve the Flexible Job-Shop Problem with recirculation. Various practical design issues are addressed in the implemented algorithm, ClonaFLEX; first, an efficient antibody representation which creates only feasible solutions and a bootstrapping antibody initialization method to reduce the search time required. Second, the assignment of suitable mutation rates for antibodies based on their affinity. To this end, a simple yet effective visual method of determining the optimal mutation value is proposed. And third, to prevent premature convergence, a novel way of using elite pools to incubate antibodies is presented. Performance results of ClonaFLEX are obtained against benchmark FJSP instances by Kacem and Brandimarte. On average, ClonaFLEX outperforms a cultural evolutionary algorithm (EA) in 7 out of 12 problem sets, equivalent results for 4 and poorer in 1. Keywords: Immune Algorithm, Clonal Selection, Flexible Job-Shop Scheduling Problem, Optimization. [ABSTRACT FROM AUTHOR]

Recent studies in classification have proposed ways of exploiting the association rule mining paradigm. These studies have performed extensive experiments to show their techniques to be both efficient and accurate. However, existing studies in this paradigm either do not provide any theoretical justification behind their approaches or assume independence between some parameters. In this work, we propose a new classifier based on association rule mining. Our classifier rests on the maximum entropy principle for its statistical basis and does not assume any independence not inferred from the given dataset. We use the classical generalized iterative scaling algorithm (GIS) to create our classification model. We show that GIS fails in some cases when itemsets are used as features and provide modifications to rectify this problem. We show that this modified GIS runs much faster than the original GIS. We also describe techniques to make GIS tractable for large feature spaces - we provide a new technique to divide a feature space into independent clusters each of which can be handled separately. Our experimental results show that our classifier is generally more accurate than the existing classification methods. [ABSTRACT FROM AUTHOR]

Corner classification (CC) is a kind of algorithms for instantly classification. The feed forward neural network trained by CC algorithm can be used validly by information retrieval, especially online information retrieval. CC4 is the fourth version of CC. In this paper, the generalized distance is defined according to the construction of weight matrix of hidden layer of CC4. With the definition of generalized distance, the principle for outputs of hidden neurons in CC4 network is given. [ABSTRACT FROM AUTHOR]

One technique that uses Wang's Recurrent Neural Networks with the "Winner Takes All" principle is presented to solve the Assignment problem. With proper choices for the parameters of the Recurrent Neural Network, this technique reveals to be efficient solving the Assignment problem in real time. In cases of multiple optimal solutions or very closer optimal solutions, the Wang's Neural Network does not converge. The proposed technique solves these types of problem. Comparisons between some traditional ways to adjust the RNN's parameters are made, and some proposals concerning to parameters with dispersion measures of the problem's cost matrix' coefficients are show. [ABSTRACT FROM AUTHOR]

According to the principles of non-self detection and negative selection in natural immune system, two generating algorithms of detector are proposed in this paper after reviewing current detector generating algorithms used in artificial immune systems. We call them as Bit Mutation Growth Detector Generating Algorithm (BMGDGA) and Arithmetical-compliment Growth Detector Generating Algorithm (AGDGA) based on their operational features. The principle and work procedure of the two detector generating algorithms are elaborated in details in the paper. For evaluation of the proposed algorithms, they are tested and verified by using different datasets, and compared to Exhaustive Detector Generating Algorithm (EDGA). It turns out that the proposed two algorithms are superior to EDGA in detection performance and computational complexities. [ABSTRACT FROM AUTHOR]

PLVIP, experimental library for parallel image processing, designed and implemented in the Image Processing Laboratory of the Institute of Computational Mathematics and Mathematical Geophysics SB RAS, is described. The library is built on the principle of vertical processing and is installed on two multiprocessor computers of the Siberian Supercomputer Center, the 32-processor Linux cluster MVS-1000/M and the 8-processor SMP server RM600-E30. Basic characteristics of the library (supported data formats, organization of computational process, and implemented subprograms) and an example of its application are considered. [ABSTRACT FROM AUTHOR]

The abstract should summarize the contents of the paper and should Steganography is a technique to hide secret information in some other data (we call it a vessel) without leaving any apparent evidence of data alteration. All of the traditional steganographic techniques have limited information-hiding capacity. This is because the principle of those techniques was either to replace a special part of the frequency components of the vessel image, or to replace all the least significant bits of a multi-valued image with the secret information. Our new steganography uses an image as the vessel data, and we embed secret information in the bit-planes of the vessel. This technique makes use of the characteristics of the human vision system whereby a human cannot perceive any shape information in a very complicated binary pattern. We can replace all of the "noise-like" regions in the bit-planes of the vessel image with secret data without deteriorating the image quality. We termed our steganography "BPCS-Steganography," which stands for Bit-Plane Complexity Segmentation Steganography. [ABSTRACT FROM AUTHOR]

Designing high performance classifiers for structured data consisting of multiple components is an important and challenging research issue in the field of machine learning. Although the main component of structured data plays an important role when designing classifiers, additional components may contain beneficial information for classification. This paper focuses on a probabilistic classifier design for multiclass classification based on the combination of main and additional components. Our formulation separately considers component generative models and constructs the classifier by combining these trained models based on the maximum entropy principle. We use naive Bayes models as the component generative models for text and link components so that we can apply our classifier design to document and web page classification problems. Our experimental results for three test collections confirmed that the proposed method effectively combined the main and additional components to improve classification performance. [ABSTRACT FROM AUTHOR]

Although many systems exist for automatic classification of faces according to their emotional expression, these systems do not explicitly estimate the strength of given expressions. This paper describes and empirically evaluates an algorithm capable of estimating the degree to which a face expresses a given emotion. The system first aligns and normalizes an input face image, then applies a filter bank of Gabor wavelets and reduces the data's dimensionality via principal components analysis. Finally, an unsupervised Fuzzy-C-Mean clustering algorithm is employed recursively on the same set of data to find the best pair of principle components from the amount of alignment of the cluster centers on a straight line. The cluster memberships are then mapped to degrees of a facial expression (i.e. less Happy, moderately happy, and very happy). In a test on 54 previously unseen happy faces., we find an orderly mapping of faces to clusters as the subject's face moves from a neutral to very happy emotional display. Similar results are observed on 78 previously unseen surprised faces. [ABSTRACT FROM AUTHOR]

This paper reconsiders the established Merkle-Damgård design principle for iterated hash functions. The internal state size w of an iterated n-bit hash function is treated as a security parameter of its own right. In a formal model, we show that increasing w quantifiably improves security against certain attacks, even if the compression function fails to be collision resistant. We propose the wide-pipe hash, internally using a w-bit compression function, and the double-pipe hash, with w=2n and an n-bit compression function used twice in parallel. Keywords: hash function, provable security, multi-collision, failure- friendliness. [ABSTRACT FROM AUTHOR]

The Differentiated Services framework allows to provide scalable network Quality of Service by aggregate scheduling. Services, like a Premium class, can be defined to offer a bounded end-to-end delay. For such services, the methodology of Network Calculus has been applied successfully in Integrated Services networks to derive upper bounds on the delay of individual flows. Recent extensions allow an application of Network Calculus even to aggregate scheduling networks. Nevertheless, computations are significantly complicated due to the multiplexing and de-multiplexing of micro-flows to aggregates. Here problems concerning the tightness of delay bounds may be encountered. A phenomenon called Pay Bursts Only Once is known to give a closer upper estimate on the delay, when an end-to-end service curve is derived prior to delay computations. Doing so accounts for bursts of the flow of interest only once end-to-end instead of at each link independently. This principle also holds in aggregate scheduling networks. However, it can be extended in that bursts of interfering flows are paid only once, too. In this paper we show the existence of such a complementing Pay Bursts Only Once phenomenon for interfering flows. We derive the end-to-end service curve for a flow of interest in an arbitrary aggregate scheduling feed forward network for rate-latency service curves, and leaky bucket constraint arrival curves, which conforms to both of the above principles. We give simulation results to show the utility of the derived forms. [ABSTRACT FROM AUTHOR]

Males may use sexual displays to signal their quality to females; the handicap principle provides a mechanism that could enforce honesty in such cases. Iwasa et al. [1] model the signalling of inherited male quality, and distinguish between three variants of the handicap principle: pure epistasis, conditional, and revealing. They argue that only the second and third will work. An evolutionary simulation is presented in which all three variants function under certain conditions; the assumptions made by Iwasa et al. are questioned. [ABSTRACT FROM AUTHOR]

A compositional proof method allows the components of a system to be specified and verified independently, instead of having to verify the entire system as a monolithic unit. This paper describes how the composition principle of Abadi and Lamport can be applied to specify and compose systems that consist of both safety and progress properties, using the HOL theorem proving system. We discuss the translation of the composition principle into HOL and the resulting proof obligations, and introduce an example system, modeled after a microkernel operating system, that we composed using the method. [ABSTRACT FROM AUTHOR]

Our general motivation is to answer the question: "What is a model of concurrent computation?". As a preliminary exercise, we study dataflow networks. We develop a very general notion of model for asynchronous networks. The "Kahn Principle", which states that a network built from functional nodes is the least fixpoint of a system of equations associated with the network, has become a benchmark for the formal study of dataflow networks. We formulate a generalized version of the Kahn Principle, which applies to a large class of non-deterministic systems, in the setting of abstract asynchronous networks; and prove that the Kahn Principle holds under certain natural assumptions on the model. We also show that a class of models, which represent networks that compute over arbitrary event structures, generalizing dataflow networks which compute over streams, satisfy these assumptions. [ABSTRACT FROM AUTHOR]

Image data are omnipresent for various applications. A considerable volume of data is produced, and we need to develop tools to efficiently retrieve relevant information. Image mining is a new and challenging research field that tries to overcome some limitations reached by content-based image retrieval. Image mining deals with making associations between images from large database and presenting a summarized view. [ABSTRACT FROM AUTHOR]

Choice of ADPCM's step-size updating factors M has a sea capacity of computing that optimizes multi-variables with complex target function. There is the effective scheme such as GA or MEA but its convergence rate becomes too slowly in the neighborhood of peak value of multi-peak function to come away from local optimization. The Clone Mind Evolution Algorithm (CMEA) that introduces the clone operator to reserve the strong component of the weak individual to next iterativeness, which effect is very obvious with testing the typical function, comes into the MEA's similartaxis operator and is used to optimize ADPCM's 8 step-size updating factors. The experiment result shows that the CMEA's SNR has been reformed average 1.03dB every generation, which is exceeding MEA's by 0.4dB, in beginning five of iterativeness and overrun the MEA's from generation 5. Furthermore, the MEA's quantity of computing is equal to CMEA's by 1.67 times and the latter is of anti-prematurely. [ABSTRACT FROM AUTHOR]

It is known that the degree distribution in the World Wide Web (WWW) obeys a power law whose degree exponent exhibits a fairly robust behaviour. The usual method, linear regression, used to construct the power law is not based on any, probably existing, intrinsic property of the WWW which it is assumed to reflect. In the present paper, statistical evidence is given to conjecture that at the heart of this robustness property lies the Golden Section. Applications of this conjecture are also presented and discussed. [ABSTRACT FROM AUTHOR]

We introduce a two-party communication complexity problem in which the probability of success by using a particular strategy allows the parties to detect with certainty whether or not some forbidden communication has taken place. We show that the probability of success is bounded by nature; any conceivable method which gives a probability of success outside these bounds is impossible. Moreover, any conceivable method to solve the problem which gives a probability success within these bounds is possible in nature. This example suggests that a suitably chosen set of communication complexity problems could be the basis of an information-theoretic axiomatization of quantum mechanics. [ABSTRACT FROM AUTHOR]

Proofs contain important mathematical knowledge and for mathematical knowledge management it is important to represent them adequately. They can be given at different levels of abstraction and writing a proof is typically a compromise between two extremes. On the one hand it should be in full detail so that it can be checked without using any intelligence, on the other hand it should be concise and informative. Making everything fully explicit is not adequate for most mathematical fields since easy parts do not need any communication. In particular in traditional proofs, computations are typically not made explicit, but a reader is expected to check them for him- or herself. Barendregt formulated a principle, the Poincaré Principle, which allows to separate reasoning and computation. However, should any computation be hidden? Or only easy computations? What is easy? How can we be sure that computations are correct? In this contribution, relevant notions are discussed and a principle is introduced which allows for checkable proofs which give a choice to see on request two different types of argument. The first type of argument states why any computation of this kind is correct. The second type states a (typically lengthy) detailed low-level proof of a trace of the computation. [ABSTRACT FROM AUTHOR]

This paper presents a novel Wiener-type recurrent neural network with the minimum description length (MDL) principle for unknown dynamic nonlinear system identification. The proposed Wiener-type recurrent network resembles the conventional Wiener model that consists of a dynamic linear subsystem cascaded with a static nonlinear subsystem. The novelties of our approach include: 1) the realization of a conventional Wiener model into a simple connectionist recurrent network whose output can be expressed by a nonlinear transformation of a linear state-space equation; 2) the state-space equation mapped from the network topology can be used to analyze the characteristics of the network using the well-developed theory of linear systems; and 3) the overall network structure can be determined by the MDL principle effectively using only the input-output measurements. Computer simulations and comparisons with some existing recurrent networks have successfully confirmed the effectiveness and superiority of the proposed Wiener-type network with the MDL principle. [ABSTRACT FROM AUTHOR]

This paper presents an investigation into why software engineers do not keep to the principle of separating business logic from display. The concept of method rationale is used to establish what is supposed to be achieved by following the principle. The resulting model is then contrasted with results from in-depth interviews with practicing engineers about what they want to achieve. The difference between what the principle advocates and what engineers consider beneficial holds the answer to why the principle of separating business logic from display is not maintained. The results suggest that many espoused benefits of the principle do not appeal to engineers in practice and the principle is tailored to make it more useful in particular contexts. Tailoring the principle also brought about other benefits, not explicated by the principle, thus reinforcing the idea that method tailoring is crucial to the successful enactment of information systems engineering methods. [ABSTRACT FROM AUTHOR]

Aiming at the problems happened in the practical application of traditional Principle Component Analysis (PCA), the concept of Relative Principle Component (RPC) and method of Relative Principle Component Analysis (RPCA) are put forward. Meanwhile, some concepts such as Relative Transform (RT), "Rotundity" Scatter and so on are introduced. The new algorithm can overcome some disadvantages of traditional PCA for compressing data when data is "Rotundity" Scatter. A simulation has been used to demonstrate the effectiveness and practicability of the algorithm proposed. The RPCs selected by RPCA are more representative, and the way to choose RPCs is more flexible, so that the application of the new algorithm will be very extensive. [ABSTRACT FROM AUTHOR]

In this paper extended Hamming distance is introduced to construct the search space. According to the features of this space, a hierarchical mutation strategy is developed for the purpose of enlarging the search area with less computation effort. A local learning principle is proposed. This principle is used to ensure that no mutation operates on the same locus of chromosomes within one generation. An evaluation method called fitness effort for calculating computational effort per increased fitness value is also given. Experimental results show that the proposed hybrid approach of hierarchical mutation and local learning can achieve better performance than traditional methods. [ABSTRACT FROM AUTHOR]

Based on the principle of Bayesian theory-based forecasting, a new forecasting model, called Bayesian Least Risk Forecasting model, is proposed in this paper. Firstly, the principle and modeling idea of Bayesian forecasting are illustrated with the explanation of the meaning of least risk forecasting. Then the advantages and learning algorithm of this model are discussed explicitly. In order to validate the prediction performance of Bayesian Least Risk Forecasting model, a simulated time series and practical data measured from some rotating machinery are used to compare the ability of prediction with classical artificial neural networks model. The results show that the bayesian model can contribute to a good accuracy of prediction. [ABSTRACT FROM AUTHOR]

A system based on vertical linear array is brought forward. The localization model is presented, and the localization principle of vertical linear array is discussed. With range, elevation angle induced from time-delay and the measurement of azimuth angle, three dimensional equation of state and equation of measurement have been established. The localization precise according to length of baseline and precise of time-delay is simulated and analyzed. According to the observability principle of nonlinear system, the observability qualifications of the system are given. [ABSTRACT FROM AUTHOR]

Preference-based search is the problem of finding an item that matches best with a user's preferences. User studies show that example-based tools for preference-based search can achieve significantly higher accuracy when they are complemented with suggestions chosen to inform users about the available choices. We discuss the problem of eliciting preferences in example-based tools and present the lookahead principle for generating suggestions. We compare two different implementations of this principle and we analyze logs of real user interactions to evaluate them. [ABSTRACT FROM AUTHOR]